Lyon, a Monday morning in January. Thomas sets his still-steaming coffee on his desk and reopens the tab he had closed the night before with relief. A competitor just shipped, almost identically, the feature his team spent four months building.
“How did they pull that off in three weeks?” he asks Sophia, his technical co-founder.
She shrugs. “We had a head start. They just watched what we were doing and rebuilt it with two extra developers.”
Thomas goes quiet for a moment. He has just learned, the hard way, the difference between having an advantage and having one that lasts.
A feature is not an unfair advantage
The useful question to ask Thomas is “what in your product could not have been copied in three weeks?” His answer is simple: nothing.
An unfair advantage is something a competitor cannot easily buy, copy, or replicate (Stratrix). Three families of unfair advantage show up most often.
The first is proprietary data: a customer history or a feedback loop that improves the product with every use, and that nobody else owns. The second is know-how or network: specialized expertise or relationships built over years within a given industry. The third is regulatory protection: a patent, a license, or an approval that blocks new entrants (Highline Beta).
Thomas had built a good feature. He had none of these three things. That is why it got copied so fast.
The moat: how an edge turns into a rampart
Warren Buffett popularized the image of the moat, the ditch around a castle that durably protects a company’s profits from competitors (Conversion Rate Experts). The unfair advantage is the initial spark. The moat is what that spark becomes once it widens and hardens over time.
According to CRV’s analysis, a true moat makes your market position harder to replicate with each passing quarter, not just at launch (CRV). That is exactly what Thomas was missing: his edge was not reinforcing itself over time, he had to keep defending it with new features.
Three types of moat show up most often in the literature on the subject.
Network effects make a product more valuable as more people use it, like a marketplace or a phone network. Switching costs make moving to a competitor too expensive, too slow, or too risky for the customer. Cost advantages come from a scale of production that lets you make things cheaper than anyone else (Financial Modeling Prep).
None of these three moats can be improvised in three weeks. That is exactly what makes them moats.
The difference between unfair advantage and moat
An unfair advantage answers the question: why you and not someone else, today? A moat answers a different question: why you and not someone else, five years from now?
The first concept measures a starting point. The second measures a trajectory. You can have an unfair advantage and never build a moat, as Thomas did, letting a real edge slip away for lack of reinforcing it. You can also build a moat without a spectacular unfair advantage at the start: plenty of companies began with nothing special and patiently built the network effects or switching costs that protect them today.
This is where the optimism bias works against founders. It describes a tendency to overestimate the likelihood of positive events and underestimate that of negative ones. Many entrepreneurs assume an advantage observed at one point in time will remain an advantage forever. On an open market, it almost never does.
Turning an unfair advantage into a moat
In my book, I cover in detail the three ways to protect an invention: patent, copyright, and trade secret (My book, chapter 3). None of these three options is a moat on its own. A patent exposes the details of your technology and does not stop you from being copied, it only gives you a basis to defend yourself in court, provided you can afford it. A trade secret costs nothing but only protects you as long as nobody discovers it or leaks it.
Building a real moat happens elsewhere: in how deeply the product embeds itself into the customer’s habits.
Here are four concrete levers, in the order I usually recommend to a founder or an innovation unit leader.
First lever: identify what, in your product, improves automatically with use. If your product does not get better from the data it accumulates, you have a user base, not a data network effect.
Second lever: map the real cost, in time and risk, for a customer who wants to leave. If that cost is close to zero, your retention rests on momentary satisfaction, not on a moat.
Third lever: measure your cost curve against your direct competitors. A cost advantage that widens with volume is one of the hardest moats to copy, because it requires an upfront investment few challengers can afford.
Fourth lever: protect what can be protected by law, patent, trademark, or trade secret, while knowing that legal protection alone is never enough. It buys time. The moat gets built with that time, not instead of it.
An exercise for your organization
Take your main product or service and rate it on a scale of 0 (no protection) to 9 (nearly impossible to copy) for each of the four levers above. Most leaders discover, like Thomas, a high score on one lever and close to zero on the other three.
That imbalance is precisely why so many brilliant innovations disappear the moment a better-funded or faster competitor enters the market.
I have seen the same pattern inside large organizations, not just startups. A business unit ships an elegant new process, gets praised in a steering committee, and loses its edge within two quarters because nothing about the process made it harder for a rival division, or a rival company, to copy the following month. Size does not grant a moat. A large company without switching costs, network effects, or a real cost curve is just a well-funded version of Thomas’s original mistake.
The opposite is also true. A small team with a genuine data loop, one that gets smarter with every customer interaction and cannot be reproduced without the same years of accumulated use, can outlast a much larger and better-funded rival. That is the practical value of separating the two concepts: an unfair advantage tells you where to start, a moat tells you what to build toward.
The essentials
Remember three things:
- An unfair advantage is a starting asset, hard to buy or copy immediately: proprietary data, know-how, regulatory protection.
- A moat is what that advantage becomes once it strengthens over time, through network effects, switching costs, or cost advantages.
- Legal protection of an invention, patent, copyright, or trade secret, buys time, it does not build the moat by itself.
Thomas eventually rebuilt his product around a piece of data his customers handed him every week that no competitor owned. Three years later, the feature copied in three weeks is gone from the market. His moat is still there.
References
- (My book) https://philippeboulanger.com/book/
- (Conversion Rate Experts) https://conversion-rate-experts.com/moats/
- (CRV) https://www.crv.com/content/what-is-a-moat
- (Stratrix) https://www.stratrix.com/strategy-lexicon/unfair-advantage
- (Highline Beta) https://www.highlinebeta.com/blog/do-you-have-an-unfair-advantage
- (Financial Modeling Prep) https://site.financialmodelingprep.com/education/other/how-an-economic-moat-provides-a-competitive-advantage
